重庆科技大学学报（自然科学版）2025，Vol.27Issue(5)：70-80,11.DOI:10.19406/j.issn.2097-4531.2025.05.009
基于CNN-BiGRU的复杂断块油藏储层连通性分析及预测
Research on the Analysis and Prediction of Reservoir Connectivity in Complex Fault-Block Oil Reservoirs Based on CNN-BiGRU
摘要
Abstract
The injection-production connectivity in the high structural of Bohai P Oilfield is relatively complex,and there is a contradiction between the production performance characteristics of oil and water wells and the distribution characteristics of static faults,leading to poor injection-production efficiency.To address this issue,a convolutional neural network(CNN)was innovatively introduced to extract local features from the historical production data of oil and water wells(such as liquid production,oil production,bottom-hole flowing pressure,water injection volume,and injection pressure),which were then used as input feature sequences for a bidirectional gated recurrent unit(BiGRU).In the BiGRU,the forward GRU captures the impact of historical production and injection data on the current state,while the reverse GRU understands the potential impact of future data on current productivity through the flow of information from the future to the past,thereby fully utilizing the historical and future production infor-mation in oilfield development.The global features extracted by BiGRU are integrated with the local features extrac-ted by CNN to form a comprehensive feature set that encompasses both the local patterns and structural information of the data,as well as the bidirectional dependencies and long-and short-term memory information of the time se-ries.This integration comprehensively reflects the interactions between production wells and injection wells.By an-alyzing these integrated features,the impact patterns of injection operations in injection wells on parameters such as liquid production and pressure in production wells can be identified,thereby predicting the connectivity relationship between injection and production.关键词
复杂断块油藏/卷积神经网络/连通关系/非均质性/断层发育/双向递归神经网络Key words
complex fault-block oil reservoir/convolutional neural network(CNN)/connectivity relationship/het-erogeneity/fault development/bidirectional gated recurrent unit(BiGRU)分类
石油、天然气工程引用本文复制引用
梁潇..基于CNN-BiGRU的复杂断块油藏储层连通性分析及预测[J].重庆科技大学学报（自然科学版）,2025,27(5):70-80,11.基金项目
国家科技攻关计划项目"渤海典型油田高含水期剩余油微观富集规律研究"(ZZKY-2022-TJ-01) （ZZKY-2022-TJ-01）
